There are no studies in the dental literature on methods to predict the service life of or periodontal response to dental restorations. The goals of this 11 year recall of bonded cast prostheses (BCPs) are two fold: The first goal of this study is to apply Weibull analysis to project the service life of dental restorations by specifically predicting the time to debond on a group of patients with previously bonded BCPs. This analysis will be based upon the 11 year clinical recall data collected over a three year period on the first 368 BCPs placed at the University of Maryland. Comparison of these debond rates to those predicted from our analysis of the 3-5 year and 5-7 year recall data will be performed employing refined methods of analysis. These techniques will be verified by analysis of data on failure rates and failure times from other studies. Secondly, this project tests the hypothesis that BCPs do not contribute to an adverse periodontal response or increased susceptibility to caries. This retrospective study will involve periodontal and prosthodontic evaluations of patient functioning with BCPs for at least 8-11 years and comparing these to our earlier data. The periodontal criteria will be applied to a variety of sites on both control and experimental teeth (with bonded retainers) and consist of: Gingival Index, Plaque Index, Probing Depths, Recession and Attachment Loss. The prosthodontic criteria will be applied to experimental teeth and include: Margin Location, Margin Contour, Embrasure Contour, Margin Integrity, and Presence of Caries. The specific aims are to: 1. Compare the actual number of debond failures to a prediction curve based on Weibull analysis. 2. Use Weibull analysis to determine if the hazard rate for debonds varies with time. 3. Incorporate patients that did not complete the study for various reasons into the data set using "deeply censored data" techniques. 4. Demonstrate the application of Weibull analysis to other clinical data (debond rates for BCPs at the University of Iowa and dental amalgam replacement rates in Denmark). 5. Confirm the hypothesis that the long-term debonds of BCPs are a result of fatigue fracture through the bonding resin. 6. Determine the long-term periodontal response to BCPs. 7. Use recently developed prosthodontic criteria to test their validity in predicting a periodontal response. 8. Determine whether BCPs cause an increased susceptibility to caries.